It's definitely better at doing so than a 100 percent full building.

A Linux-based coffee shop would use dynamic deprioritization -- each customer would be assigned a default priority upon entering the shop, and customers found to be using up too much table-time would be automatically deprioritized and bumped off of their table and into a waiting-queue if necessary, until spare tables became available again.

About as much as we needed to cover the story of irrelevant companies migrating to cloud hosting?

I've dealt with sites managing tens of thousands of hosts. It's surprisingly not *that* difficult. It's a niche community but they have the tooling to do it.

It would be *excruciating* to wrangle that many laptops with laptop style usage, but not so much for servers. Besides improved manageability, there's also the reality that no one cares too much about arbitrary box X in a sea of boxes, so no one's particularly bothered if you batch up repair actions to do once every couple of weeks or maybe once a month, leaving specific boxes either completely down or somewhat degraded for weeks unnoticed by anyone who cares.

I mean, if those numbers are accurate, it seems like more than just "playing with their stats". $2.2 mil in CapEx in a year avoids $3 mil in OpEx? For this sort of stuff, you generally would expect such a capital expense to be about 3 times higher than the operational expense, with the benefit of getting more years out of it instead of incurring it every single year. If CapEx paid off within the same year, that's a really substantive benefit.

I also find it credible, as I've dealt with businesses so enamored of 'all OpEx' they said point blank they would pay more to rent the capacity for a year than it would take to buy the same capacity permanently because CapEx is just such a no-no in accounting world when it comes to IT equipment.

And people were all weirdly vocal about going to the cloud as well.

It's a big shift, between capex and opex. There are a class of investors that like capex, particularly for longer lived companies, and others that love opex, particularly if there's a large chance the company goes bust before it could get the benefit of big upfront capex.

The "everything should be opex" crowd has been absurdly the standard voice in the market, to the point where companies are encouraged to spend more opex in a year than it would cost in capex over 3 years for the same stuff.

Well that right there would suggest they have a problem. Someone with respect for their brand hasn't considered their products since last century...

I have a panasonic plasma I've thought pretty well of...

But broadly Panasonic seems to be undifferentiated products in a lot of segments, "good enough" but not doing anything that inspires significant brand loyalty.

It isn't exactly something you have to "train up" to do. You *might* get outmaneuvered in some specific ways if you neglect to bother to use it, but I don't see it as being an active employment discriminator any more than "can you let tab completion do it's work when it can save you time or do you just type everything out manually?"

Why anyone thought that was going to be a 'skill'.

The whole point of the prompt is that it's a low skill interface. If it did require skills, then it's missing the point.

The skills to audit and do the work the LLM fails to do are non trivial, the natural language prompt piece is trivial.

AI fabricates 100% of its answers. It just so happens that many of its fabrications happen to correspond to objective reality; the remainder do not. At no time does the AI know which of its answers are "true" or "false", as it has no concept of reality and no way to find out; it has only the ability to produce output that is statistically similar to the input it was trained on.

That doesn't mean it can't be a useful tool in some circumstances, but it's critical for users to understand its limitations. Everybody wants AI to be a magic all-knowing oracle, that understands what it is talking about and never generates false information, but it's only a crude simulation of that.

If not the NOAA, then who? The NOAA is the agency that was doing it, and knew how to do it.

I don't know that you would, but some people might like to have a "heads-up-display" so that they don't have to take their phone out of their pocket every time they receive a notification; they could just glance at the text message in the corner of their vision and decide right away whether it's something worth dealing with immediately, or not. YMMV.

Do non-Libyans belong in Libya, in the middle of a civil war? Do non-Salvadorans belong in a nobody-can-ever-leave oubliette in El Salvador?

That sounds less like "returning them" anywhere, since they've never been there before in the first place, and more like a deliberately cruel and unusual punishment, intended to make an example of them in order to intimidate others.

Yes, it's good to be the Hegemon: you're expected to keep the world running in a predictable and orderly fashion, but that means you are given lots of leeway to do things your way, and as a side benefit you get to take economic advantage of the world's faith in your competent leadership, in the form of cheap capital provided to you at very low interest rates whenever you sell Treasury bonds to foreigners.

That works fine (for the USA anyway) until the day the world loses faith in the USA's ability to run things competently, and starts moving its investment dollars out from US bonds and into something else that they feel is safer (Renminbi or Euros or cryptocurrency or gold or tanks or whatever). At that point the USA loses its traditional Hegemonic privileges and will have to start playing by the same financial rules as everyone else, and US citizens aren't going to like that transition very much.

Trump appears to be doing everything a President could possibly do, to make that day arrive sooner rather than later.

That would require them to trust us not to do it again. I don't know what could induce them (or anyone) to trust the USA again after two Trump presidencies.